International Journal on Recent and Innovation Trends in Computing and Communication
Not a member yet
8613 research outputs found
Sort by
Optimized Forecasting Air Pollution Model Based On Multi-Objective Staked Feature Selection Approach Using Deep Featured Neural Classifier
Get PDFIn recent days air pollution has been an essential issue affecting the environment nature leads to various natural causes. Especially the Covid-19 pandemic period has a variation environment changes due to vehicle controls and industrial facts at regular intervals. So air pollution has different scaling factors before and after the pandemic, period produces non-scaled data features. Many methodologies provide the differential solution to analyze the air quality measurements under various conditions to make warnings to avoid air pollution. By the impact of exiting forecasting, ML approaches do not provide the accuracy in precision levels because feature dependencies are non-relevant in high dimension nature. To create the best Air quality index, we need to improve the feature analysis and classification objectives to produce higher prediction performance. This paper proposes a new forecasting model based on the Multi-objective Staked Feature Selection Approach (MoSFS) using the Deep Featured Neural Classifier (DFNC) model to predict air pollution. Initially, the Successive Feature Defect Scaling Rate (SFDSR) was carried out Auto Regressive Integrated Moving Average (ARIMA) rate for finding variation dependencies. The multi-objective relational successive feature index was scaled using the Spider Herding Algorithm (SHA) to select the features based on these variations in feature limits. Then the chosen features get activated to logical activation function with Long Short Term Memory (LSTM) and trained with a Fuzzified Convolution Neural Network (F-CNN) to predict the class by variance. This resultant factor proves the performance of RMSE values attaining the best level to forecast the features and in precision rate produce higher performance in classification accuracy compared to the other system
Secure Framework for Cyber Data Using Cryptographic and Steganographic Algorithms
Get PDFThe e-commerce industry has recently seen enormous growth on a global scale. Due to the rising popularity of online shopping, consumers, businesses, and depository financial institutions are extremely worried about debit/credit card fraud and the protection of personal information. It is essential to prevent unwanted access to and use of the information since it is disseminated over insecure channels. Cryptography and steganography are most frequently employed to avoid unauthorized access to sensitive data. However, the combined qualities of these two approaches are not secure enough in the modern world. It might lead to some vulnerability. It is possible to add additional layers of protection and achieve high levels of information security if visual cryptography is used in conjunction with the abovementioned combination. This research work proposed a multi-level information security framework for cyber data using random public key cryptography for the secret text, color image steganography for concealing secret encrypted text in the cover image, visual cryptography for slicing the cover image into two shares, and image steganography again to hide both the shares into two color images, respectively. These security processes significantly increase the confidentiality, dependability, and efficiency of secret messages. While there isn't a parameter to demonstrate the level of security achieved using cutting-edge techniques, the accuracy of the received text data is calculated in terms of MSE and correlation coefficient by comparing the sent and received text data. To evaluate the effectiveness of the suggested strategy, the time required at the transmitter and receiver ends is also calculated. The MATLAB environment is utilized in the implementation, demonstrating that the suggested system has improved robustness when considering steganalysis
Risk Prioritization using A FUZZY BASED Approach in Software Development Design Phase
Get PDFThe success of a software project's objective is directly proportional to the degree to which it satisfies all of the stakeholders' concerns regarding the project's requirements, including the budget, schedule, and overall performance. Risks can occur throughout the software development lifecycle (SDLC) phases and affect every phase. The design phase of the SDLC yields an overview of the software and can be defined as the software's blueprint. Different types of software have their own unique design phases and have different types of risks. With the high number of interacting components, complex systems have a greater propensity to be more volatile, which increases the risk. It is necessary to prioritize the risks in order of their severity levels. The issue at hand is the lack of effective methods to prioritize and mitigate the risk. Recent studies have suggested several methods for prioritizing risks, but it is clear that few of these have been implemented. These methods are overly complicated, time-consuming, prone to inconsistency, and challenging to put into practice. This paper proposes a novel Fuzzy-based approach to risk prioritization in the software design phase using MATLAB software. Fuzzy-based models have been shown to be more accurate than other techniques when using standard datasets to prioritize risks. Fuzzy-based methods that have been proposed take into account the characteristics of risks by modelling those characteristics as fuzz
AI-Based Drone System for Medical Support in Congested Areas
Get PDFServing the needs of human beings is much more important to us. So, considering how to service the product or provide medicinal support is also very important to reach the destination. While supported by the local field-side forces, it is going through delays and getting affected by the delays in medical treatment. Why this failure in medical support? To avoid this delay, I have come up with a trend in technology in real life to send things like oxygen to the nearby hospital and make it emergency support on an urgent basis. At the same time, we can look for college and industry, food, and grocery items that we can utilize for the same methods. We can make it using AI technology used in the drone system and so by using different scales of sensors and cameras for recognition for bill service for auto-detection and bill payment as well. Nowadays, most people try to use their outside food purchases quickly, so we can get it via air drone easily without delay. This process is unintentionally free and GPS-based, with an advantage system to track the location. And confirmation of customer detection can help to unlock the lock-unlock the drone's locked body, get material, and lock the door once service is received. If there is any unwanted attack, it will be updated with the tracking system. There is an anti-drone system used here. This activity can be avoided.
 
Snowball-Miner: Integration of Deep Learning for Extraction of Cyber Threat Intelligence from Dark Web
Get PDFIn Cyber threat intelligence is a crucial component in defending against cybersecurity threats. Cyber security dark web, security Blogs, Hackers’ community, news forums, Open-Source Intelligence (OSINT) are known as the harbor of illicit activities and serve as a breeding ground for cybercriminals. Extracting actionable intelligence from the dark web is challenging due to its anonymous and encrypted nature. State-of-art work proposed machine learning and deep learning approach to aggregate the dark web for cyber threat intelligence from data present in the dark web. This paper proposes, a novel approach utilizing Snowball-Miner for cyber threat intelligence discovery from the dark web. The model is trained on a diverse dataset consisting of dark web forums, hidden .onion based marketplaces and other underground platforms using Snowball-crawler. However, we have employed hybrid convolutional model CNN-LSTM and CNN-GRU adopting doc2vec word embedding to classify into four domains viz Energy Sector, Finance, Illicit Activities and illegal Services. From our experiment it emerged that, CNN-LSTM outperforms as 96.37% for classification of domain specific threat documents. Furthermore, after data preparation we implemented NLP technique and extracted the domain specific Indicator of Compromise (IoCs) using RegEx parser and Subject, Object and Verb (SOV) semantics dependency analysis. Finally, we have integrated IoCs and Threat keywords with respective domains to generate domain specific threat intelligence which enhance the quality of the domain specific CTI based on R-dimension (Relevance)
Diabetic Prediction Using Hybrid Smote-Tree Big Data Classification with Artificial Neural Network
Get PDFDiabetes is one of the worst illnesses now plaguing humanity. The condition is caused by the body's abnormal reaction to insulin, a vital hormone that transforms sugar into the energy required for the normal functioning of daily living. In addition to increasing the chance of developing kidney disease, heart disease, and retinal eye disease, nerve damage, and blood vessel damage, diabetes causes serious consequences in the body. This research provides diabetes prognosis based on hybrid SMOTE-TREE large data categorization utilizing Artificial Neural networks (ANN). Artificial Neural Networks deliver promising results for nonlinear data. Hence ANN is picked for creating the model to predict diabetes among numerous ML (machine learning) techniques. The goal is to develop a decision support system to predict and diagnose diabetes with maximum accuracy, given the parameters. The parameters are set such that the best accuracy is obtained
Enhancing Performance of Deep Learning Models for Epilepsy Seizure Detection
Get PDFEpilepsy is a neurological condition marked by recurring seizures, leading to notable effects on the well-being of individuals experiencing it. Deep learning models have shown promising results in detecting and classifying epilepsy based on electroencephalogram (EEG) data and Magnetic Resonance imaging (MRI). However, achieving high performance in epilepsy detection requires continuous efforts to enhance the accuracy and reliability of these models. This study introduces multiple approaches for improving the effectiveness of deep learning models designed for detecting epilepsy. Initially, we employ data preprocessing methods to cleanse and prepare the input data, including noise removal, data normalization, and handling missing values. Additionally, data augmentation methods, such as random rotations, translations, and scaling are employed to increase the diversity and generalizability of the training data. Secondly, various model architectures are explored to improve the model's ability to detect epilepsy. CNNs and RNNs are commonly employed, and their configurations are experimented with by adjusting the depth, and width, and adding additional layers such as residual connections or attention mechanisms. Furthermore, hyperparameter tuning techniques are employed to enhance the deep learning model's efficiency. Thoughtful choices are made regarding hyperparameters like learning rate, batch size, and regularization methods and are carefully selected through approaches like grid search or random exploration conducted to discover the best possible setup that maximizes the model's effectiveness. By implementing these strategies, the performance of deep learning models for epilepsy detection has been significantly enhanced. The improved accuracy and reliability of these models offer great potential for early detection and intervention, leading to better management and treatment outcomes for individuals living with epilepsy
Modified LMS Adaptive Algorithm for the determination of Maximum SNR using Self-Adaptation Technique of Noise Factor for Communication System
Get PDFThe most often used adaptive filter (AF) is the least-mean-square (LMS) filter. This filter has many applications in the area of communication and signal processing. System identification represents a significant use case for adaptive filters. This paper represents a new structure/model of system identification and adaptive noise cancellation (ANC) using state-of-the-art LMS-AF. The paper expressed the algorithm of updated LMS-AF. The major parts of the proposed model are- two adaptive filters, one LMS filter, one correlation function, and one auto-correlation function. The research investigation for the proposed model is based on the self-adaptation or self-learning method to obtain the maximum signal to noise ratio (SNR) based on the real-time inputs. The proposed structure/model converges towards the ideal LMS-ANC system. This paper also includes mathematical analysis, simulation, results, and discussion. The most important comparison parameters are the SNR and mean-square-error (MSE)
Novel MobileNet based Multipath Convolutional Neural Network for defect detection in fabrics
Get PDFAutomatic fabric defect detection and classification is the most important process in the textile industry to ensure the fabric quality. In the existing systems, a learning based method is used for detecting defects in plain weave fabrics. In this paper, a novel MobileNet based Multipath Convolutional Neural Network (MMPCNN) architecture is proposed for detection and classification of simple and complex patterned fabric defects. In the proposed MMPCNN architecture, MobileNet model is used in the first path. In this, Gabor filter bank is used instead of conventional filters in the first convolution layer. A simple convolutional neural network architecture with Gray Level Co-occurrence Matrix (GLCM) features as an input is used in the second path of the MMPCNN architecture. Gabor filters are more useful for analyzing the texture with different orientations and scales. Each Gabor filter parameter has its own impact on analyzing the texture and extracting the information from the texture. Therefore, in this paper, the use of Gabor filter parameters in MMPCNN architecture is analyzed. The proposed model is experimented on the TILDA textile image database and it is able to achieve 100% accuracy with reduced trainable parameters for fabric defect detection and classification
Effective Cost Reduction Usage of Infrastructure in Small Scale Sectors using Cloud Storage and Internet of Things
Get PDFPresently many small-scale information technology based private sectors are scare due to spending of high cost for maintaining of infrastructures and resources. In order to avoid to pay more money on maintaining infrastructures, there is a need of an effective price reduction method for the usage of infrastructures in various computer-based private sectors which has been achieved through integration of different online and offline cloud storages along with internet of things sensors. This method delivers huge range of infrastructures, resources and services through online and offline platforms through on-demand basics of user request. This system disables or switch off unnecessary services or resources when it is unused by people in the private sectors which has been controlled through various types of sensors. This integrated cloud storage and IoT method verifies user verification and validation along with storing all information about status of infrastructures in offline and online cloud storage environment. This paper deals dynamic group audit regulation through k-means cluster technique with cryptographic algorithm. The alert messaging and alarm indication system has been enforced in this system. Combining of internet of things and cloud storage techniques which guarantees reduction of cost for usage of infrastructures and also assures privacy, security on quality of service with maintaining of better information transmission. This high skilled method preserves the internet of things with cloud storage environment with better performance evaluation in terms of bandwidth usage and computational and communication cost.